What are the types of software development?
The main types of software development are grouped by where the software runs and what it does: web, mobile, desktop, cloud and SaaS, backend and API, embedded and IoT, database, game, and data-science and AI development, plus low-code as a way of building. Most real products combine several of these types rather than being purely one.
The types of software development are the different kinds of software teams build, sorted mostly by where the software runs (a browser, a phone, a server, a physical device) and what job it does (a customer app, a data pipeline, a game). It is a way to describe the landscape so a business can say "we need a web app with a mobile companion and a data layer" instead of just "we need software." Each type comes with its own tools, skills and constraints, which is why naming the type early makes scoping, hiring and budgeting far more concrete.
For most companies the practical question behind "what type of software development do I need?" is really "what should we build, and who will build it?" — and the answer usually spans more than one type. A single product often mixes web, backend and database work, with a mobile app or an embedded component on top. That is exactly why full-cycle custom software development services exist: to design and deliver the right combination of types as one coherent product, rather than treating each piece as a separate project. The rest of this guide walks through the main types, sets them side by side, separates them from the methodologies that are easy to confuse them with, and shows how to choose.
The main types of software development
The main types of software development are web, mobile, desktop, cloud and SaaS, backend and API, embedded and IoT, database, game, and data science and AI development — with low-code and no-code as a fast-growing way of building several of them. Below is what each type is, where it runs, and when a business reaches for it. Read them as overlapping building blocks, not walls: most products use several at once.
1. Web development
Web development builds software that runs in a browser — marketing sites, web apps, admin dashboards and most SaaS products — and it is the most common type because a web app reaches any device without an install. It splits into front-end work (what the user sees, typically with React, Vue or Angular) and back-end work (the server logic behind it). If your users can do their job in a browser tab, this is usually where you start; our explainer on web development versus software development untangles the two terms, and if you are ready to build, web application development is the service that delivers it.
2. Mobile development
Mobile development builds apps that run on smartphones and tablets, either native (Swift for iOS, Kotlin for Android) or cross-platform (Flutter, React Native) from one shared codebase. You choose mobile when the product needs the camera, GPS, push notifications, offline use or a presence in the App Store and Google Play. It is the default for consumer-facing products and a common companion to a web app. Our guide to mobile software development covers the build approaches in depth, and mobile app development is how we deliver them.
3. Desktop development
Desktop development builds software installed and run directly on a computer (Windows, macOS or Linux), and it still matters where performance, offline reliability or deep hardware access beat the convenience of a browser. Engineering tools, creative software, trading terminals and point-of-sale systems are typical desktop territory. Modern desktop apps are increasingly built with cross-platform frameworks (such as Electron or native toolkits) and often sync to the cloud, giving local reliability with online data.
4. Cloud and SaaS development
Cloud and SaaS development builds applications hosted on remote infrastructure and delivered to many customers over the internet, so users get the software without installing or maintaining it. This is the model behind most modern business software: one hosted product, continuous updates, subscription pricing and elastic scale. It leans heavily on cloud platforms and cloud-native design; our guide to cloud software development covers the architecture, and SaaS software development covers building a multi-tenant product on top of it.
5. Backend and API development
Backend development builds the server-side half of an application — the business logic, authentication, integrations and APIs that the web or mobile front end talks to — and almost every product needs it even though users never see it. This is where data is processed and stored, where third-party services are connected, and where APIs expose functionality to other apps. When people say a product is "just" a website or app, the backend is usually the larger, harder engineering effort underneath.
6. Frontend and full-stack development
Frontend development is the discipline of building the user-facing layer of web and mobile software, while full-stack development means one engineer (or team) works across both the frontend and the backend. These are less "types of product" and more "where in the stack the work sits," but they shape hiring: a small product often needs full-stack generalists, while a larger one splits frontend and backend specialists. Getting this mix right is part of designing the right software development team structure.
7. Embedded and IoT development
Embedded development builds the software that runs inside a physical device — a car, a medical monitor, a router, an appliance — on constrained hardware rather than a general-purpose computer, and IoT development adds connectivity so those devices talk to the cloud. You reach for this type only when there is hardware in the picture, and it carries the strictest timing, safety and certification demands of any category. See our guides to embedded software development and IoT software development for how the two relate.
8. Database development
Database development designs and builds the data layer that nearly every application depends on — the schema, queries, indexes and pipelines that store and serve a product's information reliably and fast. It spans relational databases (PostgreSQL, MySQL), NoSQL stores and the data-warehouse and analytics layers behind reporting. It rarely stands alone as a product, but weak data design is one of the most expensive mistakes to fix later, so it is a discipline in its own right within most builds.
9. Game development
Game development builds interactive entertainment across console, PC and mobile, and it is its own world of engines (Unity, Unreal), real-time graphics, physics and audio. It overlaps mobile and desktop development but demands specialist skills that general application teams do not carry. Businesses outside gaming still borrow its techniques for simulations, training, and interactive 3D and AR experiences.
10. Data science, AI and machine-learning development
Data science and AI development builds software that learns from data rather than following only hand-written rules — recommendation engines, forecasting, computer vision, and the language models behind today's AI features. It is a distinct type because the workflow revolves around data, models and evaluation, not just code, and it usually plugs into a product built with the other types. Our guide to machine learning software development covers how these systems are actually delivered.
11. Low-code and no-code development
Low-code and no-code development builds applications visually with minimal hand-written code, letting teams assemble internal tools, simple web apps and workflows far faster than coding from scratch. It is best understood as a way of building the types above rather than a separate type — you might build an internal dashboard on a low-code platform and the customer-facing product in custom code. Gartner expects 70% of new applications to use low-code or no-code technology by 2026, so most companies now use it for the standard parts and reserve custom engineering for what makes them different.
Types of software development at a glance
Here are the main types of software development side by side — where each runs, a typical tech stack, and the kind of product it fits — so you can match your idea to a starting point quickly. Treat "typical stack" as illustrative, not prescriptive; the right tools depend on your team and constraints.
| Type | Where it runs | Typical stack | Best for |
|---|---|---|---|
| Web | Browser, any device | React / Next.js, Node, PostgreSQL | Web apps, dashboards, most SaaS |
| Mobile | iOS & Android devices | Swift, Kotlin, Flutter, React Native | Consumer apps, on-the-go tools |
| Desktop | Installed on a computer | Electron, .NET, native toolkits | High-performance, offline tools |
| Cloud & SaaS | Hosted, over the internet | AWS/GCP/Azure, containers, APIs | Multi-customer hosted products |
| Backend & API | Servers behind the app | Node, Python, Go, Java, databases | Logic, integrations, APIs |
| Embedded & IoT | Inside physical devices | C, C++, Rust, RTOS, MQTT | Hardware and connected devices |
| Database | The data layer | PostgreSQL, MySQL, NoSQL, warehouses | Storing and serving data reliably |
| Game | Console, PC, mobile | Unity, Unreal, C++, C# | Games, simulations, 3D/AR |
| Data science & AI | In and around the product | Python, PyTorch, data pipelines | Learning from data, AI features |
| Low-code / no-code | A way of building the above | Visual platforms, connectors | Internal tools, fast prototypes |
Types vs. methodologies: what's the difference?
A type of software development is what you build and where it runs; a software development methodology is how the team builds it — and mixing up the two is one of the most common sources of confusion because both get called "software development processes." The type is the product (a web app, embedded firmware); the methodology is the way of working (Agile, Scrum, Kanban, Waterfall, DevOps). They are independent: you can build any type with any methodology, and a single company often runs the same process across web, mobile and backend work.
The distinction matters when you scope a project. Choosing a type decides the skills, stack and constraints; choosing a methodology decides how the work is planned, reviewed and shipped. A useful way to keep them straight:
- Type answers "what kind of software?" — web, mobile, desktop, cloud, embedded, data.
- Methodology answers "how will we run the project?" — Agile, Scrum, Kanban, Waterfall.
- They combine freely. You might build a mobile app (type) using Scrum (methodology), or embedded firmware (type) using a stage-gated Waterfall process (methodology).
If it is the "how" you are weighing rather than the "what," our guide to software development methodologies compares Agile, Scrum, Kanban, Waterfall and DevOps in detail, and the software development life cycle guide covers the stages every project moves through regardless of type.
Which type of software development do you need?
Choose the type of software development by where your users are and what the product must do — not by what is fashionable. The fastest way to narrow it down is a few concrete questions about the product and its users, because the answers point straight at a type (and usually at the two or three types you will need together). Work through these:
- Where will people use it? In a browser on any device → web. On a phone, with camera/GPS/offline/app-store needs → mobile. On a powerful workstation, offline → desktop.
- Is there hardware involved? If the software lives inside a physical device, it is embedded — and if that device connects to the internet, it is embedded plus IoT.
- How many customers share it? Many customers using one hosted product points to cloud and SaaS; a single organisation's internal tool may not.
- What does it do with data? Every product needs a database; if it must learn, predict or recognise, add data science and AI development.
- How fast and how custom? Standard internal tools can start low-code; differentiated, high-scale products need custom engineering where it counts.
Notice that almost every answer also implies backend and database work, because a web or mobile front end is only the visible tip of the product. The honest answer to "which type?" is usually "a primary type your users touch, plus the backend and data layer behind it" — and when the mix is genuinely unclear, a short discovery phase with an experienced team is cheaper than committing a budget to the wrong shape. That trade-off between building fast and building to last is the same one behind the choice of custom software versus off-the-shelf.
How the types combine in real products
In practice, almost no product is a single type — a typical modern application stitches several types together, which is why "full-stack" and "product engineering" describe most real builds better than any one label. A consumer fintech product, for example, is a mobile app (mobile) and a web dashboard (web), talking to server-side logic and APIs (backend), sitting on a database and analytics layer (database and data), hosted as a multi-tenant service (cloud and SaaS), maybe with an AI feature for fraud detection (data science and AI) — six types in one product. The value of naming the types is not to force a product into one box, but to make sure every layer it needs is planned, staffed and budgeted.
This is also why the team you build with matters as much as the type: a single product spanning six types needs either generalists who cover several or a coordinated set of specialists. Deciding that mix — and whether to build in-house or with a partner — is the subject of our guides to team structure and hiring a dedicated development team. The takeaway: pick the primary type your users experience, then design deliberately for the supporting types underneath it.
2026 trends reshaping the types
The types of software development are stable, but how they are built is changing fast in 2026 — AI assistance, low-code and cloud-native delivery are cutting across every type at once. These shifts do not add new categories so much as change the economics of the existing ones. The ones worth knowing:
- AI-assisted development is now the default. Around 85% of developers regularly use AI tools for coding, debugging and review in 2026, and Gartner expects 45% of new application code to be AI-generated by 2028 — a productivity shift that touches web, mobile, backend and data work alike.
- Low-code keeps expanding. Gartner expects 70% of new applications to use low-code or no-code by 2026, and forecasts that by 2028 roughly 80% of technology products will be built by people outside traditional IT — reshaping who builds the standard parts of each type.
- Cloud-native is the baseline. New web, SaaS and backend work is built cloud-first with containers and managed services, blurring the line between "web," "cloud" and "backend" development into one delivery model.
- AI moves into the product, not just the toolchain. Data science and AI development is spreading from standalone projects into features inside ordinary apps, so more products now carry a data-and-AI layer as standard.
- Edge and embedded intelligence grow. More processing runs on devices at the edge, pulling embedded and IoT development closer to data and AI work than ever before.
The practical implication for a business is that the type you choose still depends on your users and product, but the cost, speed and team shape of building it are all shifting — which makes the choice of an experienced, AI-fluent partner more important than the choice of any single technology.
FAQ
What are the different types of software development?
The main types of software development are usually grouped by where the software runs and what job it does: web development (software that runs in a browser), mobile development (apps on iOS and Android), desktop development (programs installed on a computer), cloud and SaaS development (software delivered over the internet), backend and API development (the server-side logic and data behind an app), embedded and IoT development (software inside physical devices), database development, game development, and data science, AI and machine-learning development. Low-code and no-code development is a fast-growing way of building these, not a separate category. Most real products blend several of these types rather than being purely one.
How many types of software development are there?
There is no official count — the number depends on how finely you slice the field. A practical list that covers what most businesses build is roughly ten: web, mobile, desktop, cloud and SaaS, backend and API, embedded and IoT, database, game, and data science and AI development, plus low-code and no-code as a cross-cutting way of building. You will see articles quote anything from 6 to 40-plus types because some split frontend from backend, or list every specialism (security, DevOps, blockchain) separately. The exact number matters far less than picking the type that matches where your users are and what your product has to do.
What is the most common type of software development?
Web development is the most common type of software development, because almost every business needs something that runs in a browser — a website, a web app, an admin dashboard or a SaaS product — and web apps reach any device without an install. Mobile development is a close second for consumer-facing products, and backend development underpins nearly all of them, since a web or mobile front end almost always needs server-side logic, APIs and a database behind it. In practice most projects combine web or mobile on the front with backend and database work behind.
What is the difference between types of software development and software development methodologies?
A type of software development describes what you are building and where it runs — a web app, a mobile app, embedded firmware — while a software development methodology describes how the team works to build it, such as Agile, Scrum, Kanban or Waterfall. They are independent choices: you can build any type of software using any methodology. Confusing the two is common because both are called "software development processes," but the type answers "what kind of software?" and the methodology answers "how will we run the project?"
Which type of software development should I choose for my product?
Choose the type by where your users are and what the product must do, not by what is trendy. If people will use it in a browser on any device, start with web development; if it needs the camera, GPS, offline use or app-store presence, choose mobile development; if it lives inside a physical device, that is embedded development; if many customers will use one hosted product, that is cloud and SaaS. Almost every option also needs backend and database work behind it. When you are unsure, a short discovery phase with an experienced team maps the right mix before you commit a budget.
Is low-code or no-code a separate type of software development?
Low-code and no-code are ways of building software rather than a separate type — you can build a web app, an internal tool or a simple mobile app on a low-code platform. They speed up delivery for standard applications and let non-developers assemble tools, which is why Gartner expects 70% of new applications to use low-code or no-code technology by 2026. The trade-off is less control, harder customisation and potential lock-in, so teams typically use low-code for internal tools and fast prototypes and custom code for the differentiated, high-scale parts of a product.
Last updated 16 August 2026. Adoption figures reflect widely reported 2026 industry data (low-code and AI-in-development forecasts from Gartner; AI-tool usage from 2026 developer surveys) and vary by source and definition. The type lists here are a practical grouping, not an official taxonomy — treat this as orientation for scoping, and confirm the right mix for your product with an engineer.
